Surprisingly, in this day of remarkable technological innovation, the problem of tightening a bolt so as to create in the bolt a predetermined level of tension is a problem for which there is no simple and inexpensive general solution.
The difficulty in finding a solution stems from two sources. First, while it is technically possible to make a direct measurement of the tension which a bolt carries, all such systems devised thus far are either too expensive or too delicate for widespead use. Some success has been had withe fasteners designed to include special means which give some sort of indication when a particular level of tension is reached as a bolt is tightened. My U.S. Pat. No. 3,431,812 and 3,757,630 describe such a capability.
The second source of difficulty is the role which friction plays in threaded fasteners. As matters work out in practice, when wrenching torque is used to produce tension in a bolt only about 10 percent of the torque goes into producing tension in the bolt, the remainder going to overcome friction, about up to 50 percent to friction of the nut and bolt head bearing surfaces and up to about 40 percent to friction of the threads. This is not all bad because the helical thread is essentially an inclined plane and without friction the nut would rotate and unwind as soon as the tightening wrench is removed. What is bad about friction is that its magnitude varies considerably due to a number of causes which are difficult to control. Since the friction is difficult to control it is difficult to use the magnitude of wrenching torque as a means for reliably producing accurate bolt tension of predetermined magnitude. This problem of friction variability has been circumvented in some few designs, including those described in my patents cited above, by isolating one element in the bolt head or nut so that it is loaded only by the bolt force and undergoes observable plastic deformation when this bolt force reaches a predetermined magnitude.
There is increasing pressure to develop more accurate tensioning methods because, among other reasons, greater tensioning accuracy will reduce bolting costs by allowing the specification of higher bolt tensions, tensions close to the proof load (which is the tensile force which will produce in a bolt a specified amount of permanent elongation). Indeed, there are many users of bolts who feel that, for most applications, the solution to the problem of producing accurate and uniform bolt tensions is to tighten each bolt until its yield point is just exceeded. The reasoning is that this yield point is an inherent property of the bolt material and, hence, if a number of bolts made of the same material are tightened under varying conditions of friction until they yield, then each will carry the same tension; namely the tension which causes yielding of the bolt material. For example, two wrenching systems for tightening bolts in this way have been announced in the technical press recently (Machine Design, Volume 47, No. 2, Jan. 23, 1975, and Design News, Volume 30, No. 17, Sept. 8, 1975, page 57, both incorporated by reference herein). Both systems incorporate means for measuring the wrenching torque and the rotation of the nut (or bolt) and for determining the slope of the torque-rotation curve by calculating the ratio of increments in these two measured quantities. As the bolt tension reaches and exceeds the yield point there is a gradual, but substantial, drop in this slope as the bolt material passes through the transition zone between elastic and plastic deformation. When this drop in slope reaches a predetermined value electronic circuits are triggered to stop the tightening. Other circuits then are actuated to check whether the final values of wrenching torque and nut rotation fall within predetermined limits which determine acceptability.
Too low a final wrenching torque will indicate that the bolt's yield point and, hence, final tension lies below specification, but this is not certain, since the low torque also could result from the friction being low for that particular bolt. Too high a torque will indicate that the bolt's yield point and, hence, final tension lies above specification, but again this is not certain, since the high torque also could result from the friction being high for that particular bolt. Too small a final nut rotation will mean there has been little plastic deformation and will imply that the bolt has a high hardness and hence may be brittle. Too large a nut rotation will mean there has been substantial plastic deformation and will imply that the bolt has a low hardness and hence has a low tensile strength, and may undergo excessive plastic elongation if, under service conditions, the bolt is subjected to additional increments of tension.
Such wrenching systems as described will produce uniform bolt tensions only to the extent that the bolts, as manufactured, have the same force-deformation curves beyond the elastic range. When the bolts have different force-deformation curves, which depend greatly on the hardness level to which the individual bolt is hardened, the bolt tensions produced by these wrenching systems will vary and some of the bolts will have to be removed and discarded, an expensive method of quality control, since it includes the cost of installing and removing each defective bolt. While it is possible to control material properties and heat treatment so that all bolts, as manufactured, have substantially the same force-deformation curve beyond the elastic range, the costs of maintaining such close controls would be prohibitive. To make manufacture economical, the specifications for bolts (see Metals Handbook, Volume I, Properties and Seclection of Metals, 8th Edition, 1961, page 175, Table 2) give a range of hardness within which all bolts of a given grade and size must lie, and specify as the minimum values for the proof load stress and the tensile strength the values which will be possessed by the bolt having the minimum allowable hardness. That bolts of a given grade and size, as presently manufactured and sold, do have appreciable variations is attested to by measurements of hardness and tensile strength of bolts obtained from different suppliers of the same grade and size of bolt (see Metals Handbook, op.cit., page 178, FIG. 8).
Thus, it may be seen that the problem of securing uniform bolt tensions is not solved by tightening each bolt until its yield point is just exceeded, even with the assistance of wrenching systems such as those described above, because of the inherent variability in the yield points of bolts as presently manufactured.